Published in Hydropower & Dams Issue Three, 2017

The design of unlined hydropower tunnels and shafts:

100 years of Norwegian experience
A. Palmström, RockMass Company, Norway
E. Broch, Norwegian University of Science and Technology, Norway
The main criteria for the successful design of the unlined pressure tunnels described here are based on 100
years of experience and development of more than 80 unlined pressure waterways in Norway, with maximum
heads up to 1000 m. International experience of collapses in special, challenging ground conditions from other
countries are also included. Recommendations that have emerged from this review are to locate tunnels in
suitable rocks with sufficient confinement to avoid hydro-jacking, which could lead to water leakage failures,
and, to detect and provide sufficient support to local sections with unstable rockmass conditions, including
swelling and/or friable materials.

The term unlined tunnel is used to describe hydropower tunnels where water is in direct contact with the rock, that
means, only limited parts of the tunnel are lined with concrete or shotcrete to protect against local tunnel collapses
or major rock falls.
The benefit of using unlined tunnels and/or shafts is a significant reduction in construction costs and
construction time. Unlined tunnels and shafts with heads of more than 1000 m function satisfactorily. This has
been made possible by 100 years of development in design, construction and operation, and Norwegian experience
has made an important contribution. Today’s design, construction and operation of unlined pressure tunnels and
shafts are discussed in this paper.

1 Norwegian experience in the application of unlined waterways

1.1 Development of hydro plant layouts

Fig. 1. Trends in the developments of the general layout of
hydropower plants in Norway [Broch, 19821].
Before 1950, a surface powerhouse with a penstock was the
conventional arrangement for hydropower plants, as shown in
Fig. 1. Early powerplants used either a steel penstock from the
reservoir, a river intake or a short, unlined headrace tunnel to
the power station. Depending on the topography, the penstock
was installed at the surface or in a tunnel.
During and shortly after the First World War (1914-18)
there was a shortage of steel, which led to uncertain delivery
and very high prices. As a result of this, four Norwegian
hydropower stations with unlined pressure shafts were put
into operation during the years 1919 to 1921. The maximum
heads varied from 72 to 152 m. Although three unlined
pressure shafts, constructed around 1920, were operating
without problems after some initial problems had been solved,
it took almost 40 years for the record of 152 m of water head
in unlined rock to be beaten. By 1958, nine more unlined
pressure shafts had been constructed, but all with water heads
of less than 100 m.
As Fig. 2 shows, new unlined pressure shafts were
constructed in the early 1960s, and since 1965 unlined
pressure shafts and tunnels have been the conventional
solution. Today, more than 80 unlined high-pressure shafts or
tunnels with heads of more than 150 m are operating
successfully in Norway. The highest ‘unlined’ water head is
more than 1000 m.

1.2 Development of design

The first useful experience of unlined pressure tunnels was the construction of the Herlandsfoss hydro plant, see
Fig. 3, where a 150 m-long unlined pressure tunnel was built with a water head of 123 m. During the first filling of
the shaft and the tunnel, increasing leakage through a mica-schist layer was observed A. The tunnel was then
dewatered and the penstock was extended through the tunnel to the foot of the shaft, as the Figure shows. No leakage

A
This can easily be explained, as the shortest distance from the cone to the surface is only 17 m, which compared with the required 55 m,
is far too short (see Fig. 3).

www.RockMass.net
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

from the shaft was observed subsequently, and the powerplant has now been in operation for almost 100 years
[Selmer-Olsen, 19703; Broch, 19824 and 19845].

Fig. 2.
The development of unlined
pressure shafts and tunnels in
Norway [updated from Broch,
19845, 20102].

1.2.1 Design by ‘rule-of-thumb’

No calculations were used to locate the
unlined tunnels before 1970. Then a simple
‘rule-of-thumb’ was introduced in the
planning of unlined pressure shafts. It was
associated with the general layout for
hydropower plants used at that time (Fig. 1).
The rule-of-thumb was based on the
condition that the tunnel should be located
deep enough, so that the internal water
Fig. 3. Herlandsfoss hydropower plant, in operation since 1919. [revised from pressure was balanced by the weight of the
Vogt, 19229]. overlying rock. This simplified approach
was generally adequate where the
overburden surfaces were relatively
horizontal, and the rock stress situation
normal.
In 1968, the unlined pressure shaft with
Fig. 4. a 300 m water head at the Byrte hydropower
The conditions at plant failed and flooded the underground
Askora hydro- powerhouse, [Brekke et al., 19706]. Two
power plant. years later another less serious failure
[revised from
occurred at the Askora hydropower plant,
Berg- Christensen,
19757]. where an unlined tunnel in quartzitic rocks
with a head of approximately 200 m was
hydraulically split, see Fig. 4. The tunnel
split with leakage at Askora followed sand-
filled, steeply dipping joints with a strike
parallel to the very steep valley side (55º)
and normal to the tunnel. The failure is
described in detail by Bergh-Christensen
Fig. 5. [19757]. The remedial measure was to move
Definitions for the penstock cone 320 m upstream. The
the new rule of plant has been operating successfully since
thumb for
then. Panthi and Basnet [20168] present
unlined, high-
pressure tunnels more details about failures at Norwegian
or shafts [Broch, hydropower projects.
19824].

1.2.2 Updated rule-of-thumb

After these failures, a revision of the rule-of-thumb was introduced by Bergh-Christensen and Dannevig [197112],
whereby the inclination of the valley side was taken directly into account, as well as the density of the water and
the rock, see Fig. 5.
Based on this formula, a diagram showing existing unlined pressure shafts with or without leakage was presented.
This was supplemented with unpublished information from the Norwegian Geotechnical Institute (NGI, 197210) and
is shown in a slightly revised version in Fig. 6, with later unlined tunnels and shafts.

Published in Hydropower & Dams. Issue Three, 2017

2
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

Fig. 6. 1.2.3 Evaluation of the topography in the design

Unlined Careful evaluation of the topography in the
pressure vicinity of the pressure tunnel or shaft is neces-
shafts in
sary to assess the distance to the terrain surface.
valley sides
with various This is particularly important in areas where
valley slopes streams and creeks have eroded deep and
[revised from irregular gullies and ravines in the valley sides.
NGI, 197210 The remaining ridges, or so-called noses,
and Broch, between such deep ravines will, to a large extent
19824]
be stress-relieved. They should therefore not be
taken into account when the necessary over-
burden for unlined pressure shafts or tunnels is
measured. This does not mean that pressure
tunnels should not be running under ridges or
noses, but only that the extra overburden this may
give, should not be accounted for in the design
unless the stress field is verified through in-situ
measurements. An example is described in some
detail by Broch [198413] in relation to a project in
Colombia. Fig. 7 shows how the ridge/nose is cut
away by changing the contour lines in the map
and subsequently revising the vertical profile
along the tunnels and shaft. See also Fig. 12.

1.2.4 Unexpected low rock stresses

Fig. 7. The Chivor hydropower plant in Colombia, which has been in
operation since 1982 [Broch, 19848].
Fig. 8. The conditions at the Fossmark hydropower plant [revised from
Garshol and Blindheim, 198811]
It is worth noting that the majority of unlined
pressure shafts, at that time where leakage has
occurred, are plotting below the curves defined
by the new rule of thumb in Fig. 6. The excep-
tions are the Bjerka (1971) and Fossmark (1986)
powerplants. At Bjerka, the penstock cone was
located in a mica schist containing minor beds of
karstic marble. The large leakage after watering
up occurred along these karstic channels and was
not caused by insufficient confinement. The cone
(the unlined end of the tunnel) was therefore
moved 90 m upstream. Since then the Bjerka
plant has been working satisfactorily.
The Fossmark plant (see Fig. 8) had been
modernized with a new unlined vertical shaft and
a pressure tunnel. The rule-of-thumb showed that
the chosen design was applicable according to the
rule of thumb, see Fig. 6
However, the tunnel crosses a number of
vertical joints parallel to the valley side from
which many water inflows occurred during
excavation of the tunnel and shaft. Before
watering up, they were sealed by grouting.
During watering, however, hydro-jacking caused
the opening of some joints, leading to
unacceptable water losses. Therefore, after some
tentative trials, the pressure tunnel and the shaft
were steel lined. The cause of failure was a
combination of unfavourable geological features
and unexpected low stresses.

The failures at Fossmark, and to some degree, also at Askora, showed that the rule-of-thumb was not sufficient
for the design. The lowest principal stresses can sometimes be different from the estimated one. In subsequent years,
the common practice has been to take stress measurements close to the location of the penstock cone, to ensure that
the rock stresses are sufficient. If not, a new location for the cones is found after stress measurements.
The failure at Bjornstokk plant in 2016 took place as a result of unexpected, low vertical stresses (see Fig. 9).
This might have been avoided, had stress measurements been taken. The failure caused two landslides of 5000 m3

Published in Hydropower & Dams. Issue Three, 2017

3
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

and 130 000 m3, at distances of 400 m and 600 m from the tunnel.
Both the rule-of-thumb (in Fig. 5) and the diagram in Fig. 6 are still being used for quick, preliminary estimations.

1.2.5 Design charts based on finite element models

In parallel with the revisions to the rule-of-thumb, the search for better and more general design criteria was
intensified at the Department of Geology at the Norwegian University of Science and Technology.
A new design tool was adopted in 1971-72. It is described in detail by Brekke et al. 19706] and by Broch [19824].
It is based on the use of computerized finite element models (FEM) and the concept that nowhere along an unlined
pressure shaft or tunnel should the internal water pressure exceed the minor principal stress in the surrounding rock
mass. The charts have been of great help in early phases of planning.
Later, modern commercial and more advanced computer programs were developed and used.

1.3 Controlled watering

A further development in the design and use of unlined pressure shafts and tunnels was to conduct controlled initial
watering of the tunnel, and to carry out leakage measurements to detect any unforeseen leakage. With this procedure,
the unlined tunnel or shaft is watered in stages. After infilling over a period of 10 to 30 hours, the water level in the
shaft or tunnel is continuously and accurately monitored during a pause of 10 to 20 hours by an extra-sensitive
manometer before the next step in the watering up. Any serious leakage can be discovered early by this, and
necessary actions taken in due time. If not arrested, such seepage could lead to landslides like those at the Bjornstokk
hydro plant (see Fig. 9).
Table 1 shows the infilling and dewatering rates used or Norwegian unlined tunnels.

Fig. 9. Layout and

section of the Bjornstokk
hydropower plant,
showing the probable
failure principle. The
leakage water moved
more than 500 m and
initiated two landslides.
[Palmström and Buen,
201714].

Table 1: Rates of tunnel watering up and dewatering used in 1.4 Tunnel support
many Norwegian unlined power tunnels and shafts
The main purpose of rock support is to reinforce unstable
Operation Increase/decrease in head ground conditions in the tunnel during excavation, to obtain
per hour per day safe working conditions for the crew (often referred to as
First watering-up 10 m(a) max. 200 m initial or temporary support) and a stable tunnel during the
Later watering-up 15 m(b) max. 300 m tunnel’s operating life (called final or permanent support).
Dewatering 10–15 m (c)
max. 250 m Appropriate rock support in a tunnel may be chosen and
tailored according to the ground conditions encountered
(a)
Stop infilling, min. 2 hours for observations and measurements per
100 m head increase (10kg/cm2).
during construction. This requires follow-up of the
(b)
Stop infilling, min. 2 hours per 150 m head increase. excavation works by experienced engineering geologists.
(c)
Some stops during dewatering to be evaluated for each tunnel. In Norwegian unlined tunnels, the underlying design
philosophy for operation of the plant is that minor rockfalls can be tolerated. As long as rockfalls in local parts of
the tunnel do not develop significantly and increase the head loss, a reasonable number of minor rock falls spread
out along the tunnel will not harm it or disturb operation of the hydro station. If necessary, rockfalls may be removed
during later inspection and maintenance.
Normally a rock trap is located at the downstream end of the headrace tunnel to prevent rock fragments entering
the turbine.

2 Modern design of unlined hydro waterways

A detailed, initial geological survey is essential in the planning of hydropower tunnels. In addition to assessing the
geological features of the site, the survey should aim to identify possible difficult rockmass conditions, such as
major faults and unsuitable rocks, or other features of importance likely to intersect the tunnel (see section 2.2).

Published in Hydropower & Dams. Issue Three, 2017

4
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

The two main challenges in the design of

unlined pressure tunnels and shafts are to:
• avoid hydro-jacking resulting in major
leakage, which means locating the tunnel
with sufficient rock stresses (confine-ment);
and,
• avoid a collapse blocking the water flow in
the tunnel during power production.

The aim of the design described next is to

prevent those two types of accidents, and based
on this and other requirements, to arrive at a
successful result.
Fig. 10. An unlined headrace tunnel with an unlined pressure shaft and
unlined pressure tunnel.
2.1 Modern layout
Fig. 10 shows simplified examples of the
Fig. 11.
Example of a application of modern, unlined waterways in
plan and cross two vertical sections. Fig. 11 presents the plan
section of an and cross section of a modern underground
underground powerhouse equipped with one turbine. Similar
hydro layouts can be found at Norwegian plants with
powerhouse
with an unlined
heads in the range of 200 to 600 m. The Tjodan
tunnel and shaft plant, described by Palmström and Schanche
[Broch, 19824] [198715] is an example of a design for a 950 m
water head on unlined rock.
The layout should include future access
ways to the tunnel and shaft for inspections,
which are large enough for equipment to handle
maintenance and repairs if a collapse should
take place.
Fig. 12. The critical point to avoid potential leakage
The Nore failure is to locate the tunnel or shaft with
hydropower sufficient confinement, that means, ensure that
plant, the minor rock stresses are higher than the water
refurbished in
head at any point in the tunnel or shaft. This is
1995 [revised
from Hope et normally where the unlined pressure shaft or
al., 199719]. tunnel ends, and the steel lining or pen- stock
starts (at the penstock cone). The length of the
steel lining to the underground powerhouse is
commonly in the range of 30 to 80 m,
depending on the head and the rockmass
conditions. The access tunnel to the penstock
cone, that is, to the start of the unlined pressure
tunnel or shaft, is plugged with a concrete plug
with an access gate. The steel tube (cone) with
a hatch cover is in another concrete plug, see
Fig. 11. The length of each of the two plugs is
normally 10 to 40 m, depending on the head and
the geological conditions. As a general rule, the
length of the concrete plug is 4 per cent of the
water head, which theoretically gives a
maximum hydraulic gradient of 25. Around the concrete plugs, thorough high-pressure grouting of the rocks as well
as contact grouting are carried out. This avoids leakage into the powerhouse and the access tunnel. Further details
about the design of high-pressure concrete plugs are given in Dahloe et al. [199216] and in Broch [199917].
Normally the unlined headrace tunnel is designed with a water flow of approximately 1 m/s. The various types
of permanent support will have different roughnesses. The head loss also varies with the tunnel size. The Manning’s
formula or similar is used to calculate the head loss, with input for the roughness of the tunnel’s surface. The
Manning’s formula can be used to compare head loss in tunnels with different support, given as:
A1/A2 = (M2/M1)0.75
where A = area (cross section); and, M = Manning’s number, (see Table 2).

Published in Hydropower & Dams. Issue Three, 2017

5
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

Table 2: As an example: For an unlined tunnel excavated by drill and

Typical Manning's number for various tunnel surfaces blast with Manning’s number M1 = 35, and another concrete
Type of tunnel surface Manning’s number lined tunnel with M2 = 65, the unlined tunnel must be 1.6 times
Unlined rock (drill and blast) 25 - 40 larger than the concrete lined tunnel to have the same head loss.
Shotcrete lining 45 - 55 Shotcrete will level out some of the irregularities in a drill-and-
Concrete lining 62 - 83 blast excavated tunnel, increasing the Manning’s number. Using
Steel lining 71 - 100 M1 = 50 (and M2 = 65), the shotcrete-lined tunnel must be 1.2
times larger than the concrete-lined tunnel to have the same head loss.

2.2 Location of the unlined pressure tunnel or shaft

2.2.1 Suitable rocks
Rocks suitable for an unlined water tunnel are:
• Crystalline rocks;
• unaltered igneous rocks;
• volcanic rocks; and,
• metamorphic rocks, as well as many sedimentary rocks.

Less suitable rocks are:

• some loose, friable and/or porous rocks;
• rocks containing swelling and/or slaking minerals B; and,
• some karstic rocks.
The unlined solution should not be used where such conditions occur along a significant part of the tunnel length.
Crossing weakness zones/faults with slaking or swelling materials should be avoided. If this is not possible, careful
sealing and grouting should be carried out, in addition to appropriate rock support. Special treatment of swelling
and slaking rocks is a prerequisite. This has been described by Palmström and Stille [201518].

2.2.2 Influence of rock stresses

The entire tunnel, including the shaft and the surge shaft, must be set deeply enough within the rock mass to ensure
that sufficient in-situ compressive rock stress is available to prevent hydraulic splitting or jacking.
A first evaluation of the location of the unlined pressure shaft or tunnel normally involves a careful examination
of the topography based on a general rule that the assumed vertical stresses calculated from the overburden are
higher than the water head in the tunnel or shaft (see section 1.2.2 and Fig. 7). Ridges, or so-called ‘noses’ between
deep ravines or depressions will to a large extent be stress-relieved. They should therefore not be taken into account
in the evaluation of overburden for unlined pressure shafts or tunnels [Broch, 198413]. The Nore hydropower plant
is an example, shown in Fig. 12, describing how the ridge/nose is cut away on the map by changing the contour
lines in the map and subsequently revising the vertical profile along the tunnels and shaft. Based on this, a
preliminary location of the unlined tunnel and penstock cone was chosen. Further details about the Nore plant are
given in Hope et al. [199719]. This topographical exercise is also useful in the input to numerical analyses.
In many cases, reliable stress measurements are difficult to take before construction begins, because the critical
point, that is, the penstock cone, is located deep below the surface. This was the case for the Nore plant. Therefore,
the construction contract had some flexibility to locate the penstock cone according to the results of stress
measurements performed during excavation of the access tunnel (see Fig. 12). The first stress measurements were
taken after half of the access tunnel had been excavated. They showed stress magnitudes as expected. A second
measurement was taken 25 m before the access tunnel had been driven to the planned rock trap. As this also showed
acceptable stress magnitudes, the cone could be located as planned.

2.3 Rock stress measurements

Rock stresses can be found either by stress measurements or by hydraulic jacking tests. This is of particular
importance in cases when the stress situation is not well known. The tests are normally carried out during the
excavation of the access tunnel, as have been described for the Nore hydropower plant (Fig. 12). To make sure that
all possible joint sets are tested in the jacking test, the test holes are normally drilled in three directions. During
testing, the water pressure in the holes is raised to a level which is 20 to 30 per cent higher than the water head just
at the penstock cone. If the testing shows that jacking of the joints may occur, the unlined part of the waterway will
have to be located deeper into the rock. This will normally mean that the whole powerhouse complex is moved
further in. A flexible contract which allows for such changes, is therefore vital when unlined high pressure shafts
and tunnels are planned. Locating the powerhouse complex deeper into the rock adds length to the access tunnel,
but not to the total waterway.

B
As described in the examples in Section 3, some tuffs have been shown to have swelling properties; swelling that is not easily discovered
by normal testing procedures.

Published in Hydropower & Dams. Issue Three, 2017

6
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

2.4 Tunnel mapping

For tunnel support designs based on observations, a prerequisite is to understand the geological conditions, the rock
mass characteristics and the ground behaviour, so as to identify appropriately and select the correct rock mass
parameter values. It is very important to be aware of, or to detect, the occurrence of some special rocks described
in Section 2.2.1, to be able to select the appropriate design and rock support. The quality of the rock supporting
works installed is essential for a tunnel to function properly during the lifetime of the powerplant.

2.5 Watering and dewatering

Fig. 13. An example of controlled, slow filling of a pressure conduit
[Buen and Palmström, 198220].
Fig. 14. Measured net water leakage from various unlined high pressure
shafts and tunnels [Palmström, 198721]
During and after the excavation of the tunnel, the
surrounding rock mass will have gradually been
drained. The first watering of a pressure shaft or
tunnel should therefore be done in a controlled way,
with leakage being measured during the process to
detect any unforeseen leakage, so that appropriate
action can be taken in time.
A normal procedure is to fill the unlined tunnel or
shaft in steps or intervals of 10 to 30 hours, as shown
in Fig. 13. During these intervals, the water level in
the shaft is continuously and accurately monitored by
an extra-sensitive manometer. By deducting for the
inflow of natural groundwater and the measured
leakage through the concrete plug, the net leakage
out of the unlined pressure tunnel or shaft to the
surrounding rock mass can be calculated. Some
typical leakage curves are shown in Fig. 14. Based
on the measurements, an average permeability
coefficient of 1 - 10 × 10-9 m/s has been calculated,
giving a leakage of 0.5 - 5 l/s per km [Palmström,
198721]
If serious hydrofracturing or hydro jacking takes
place in the tunnel during the watering process, loose
materials on the surface above can become more
saturated, and if not arrested, this leakage could lead
to landslides. An example of this is described in
Section 1.2.3 and Fig. 9.
Table 1 shows some watering and dewatering
rates used in Norway.

2.6 Follow up during power production

During the first months of operation, it is very
important to keep a continuous watch on the potential
loss of water head in the headrace tunnel and shaft.
Even small head loss increases of a few centimetres
could indicate serious fallouts of rock masses. A head
loss of as much as 1 m may indicate a serious
collapse. Then it is even more important to dewater
as early as possible.
For all unlined tunnels, the most critical test for
potential stability problems is the first dewatering.
Fig. 15. The situation at the collapse in the unlined headrace tunnel at This is an important part of the whole construction
the Nye Vinstra hydro plant.
process and should therefore not be delayed. Only
after this check can the tunnel be considered to be complete from a contractual perspective.
With respect to power generation, the dewatering should be carried out after approximately one year, according
to a pre-decided schedule. If the measurement/control of water head indicates even small stability problems, the
tunnel should be dewatered as soon as possible, regardless of the pre-decided schedule, to prevent the problems
from developing further. It is much cheaper and faster to solve a small problem than a full scale tunnel collapse.
Later tunnel inspections, after dewatering, may be carried out two to five years after the first inspection, the time
for inspection depending on the conditions observed during the previous inspection.

Published in Hydropower & Dams. Issue Three, 2017

7
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

3 Examples of collapses in unlined headrace tunnels during power production

3.1 The Vinstra hydro plant
The unlined pressure tunnel, in phyllite, at this Norwegian plant, was watered in 19 9. Two years after a collapse
occurred, involving a volume of 20 000 m3 which extended for a length of 2 km downstream of the collapse (see
Fig. 15). Adequate rock support of a sub-vertical weakness zone had not been provided in the invert. The collapse
would have been avoided if the rock support works had been closely followed up, or the problem might have been
discovered if inspection had been done earlier. The remedial works took only four months, thanks to good planning
and an effective contractor. Easy access to the tunnel with modern equipment during the remedial works was
important for efficient execution.

3.2 La Higuera hydro plant

At this plant, in Chile, a collapse of the headrace tunnel occurred in 2011. An increase of head loss was registered
after some months of power production, but it took one year before the tunnel was dewatered and inspection could
take place to find a collapse of 12 000 m3, with debris extending 500 m downstream. The drill-and-blast excavation
in volcaniclastic rocks (tuffs) had passed the potential collapse area without it being observed that the fault there
had swelling potential. One reason for this was that the contractor applied shotcrete on the tunnel periphery
immediately after the blast. This made it very difficult for the engineering geologist to study how the rockmass
behaved over time. Later investigations found that the slow swelling of minerals was associated with frequent zeolite
veins in the fault. If the tunnel had been dewatered earlier, the extent of the collapse could have been reduced
considerably. Had preliminary watering been done, as was the case for the tunnel at La Confluencia hydro plant,
the behaviour of the fault could have been discovered. Better access to the headrace tunnel to carry out repair works
could have reduced cost and time for the remedial works, which took two years and required a 238 m-long bypass
tunnel.

3.3 La Confluencia hydro plant

This is located upstream of La Higuera plant in similar rock conditions. It was constructed at the same time as the
collapse occurred at La Higuera. By taking advantage of the experience with swelling rocks at La Higuera, a test
for rapid checking of the swelling using ethylene glycol was developed [Carter et al., 201022]. To check the site
conditions further, the tunnel was watered, and after approximately one month, dewatered to inspect the ground
behaviour. Additional support was installed where swelling had been occurring, and the powerplant was
successfully put into operation, and has been working without problems.

3.4 Glendoe hydro plant

This plant, in Scotland, is located in phyllites. The headrace tunnel was excavated by TBM. After only six months
of power production, the first minor increase in head loss was registered. The head loss systematically increased,
but it took four months before the tunnel was dewatered for inspection. A huge collapse of 20 000 m3 was
discovered. The contractor was fully aware of a potential fault zone in the area of the collapse. In spite of careful
observation during the TBM excavation, no clear indication of a potential collapse of the zone was observed by the
two teams of engineering geologists who mapped the tunnel. If the tunnel had been dewatered and inspected soon
after the first head loss was registered, the extent of the collapse could have been significantly reduced. The remedial
works, involving a 606 m-long bypass tunnel, took three years. This time could have been reduced, if the powerplant
had been designed with a permanent access to the headrace tunnel.

References
1. Broch, E., “Designing and excavating underground powerplants”; Water Power & Dam Construction Vol.
34, No.4, 1982.
2. Broch, E., “Tunnels and underground works for hydropower projects, - lessons learned in home country and
from projects worldwide”. Muir Wood Lecture 2010. International Tunnelling and Underground Space
Technology 2010.
3. Selmer-Olsen, R., “Experience with unlined pressure shafts in Norway”. In: Large permanent underground
openings, Oslo, Norway. Universitetsforlaget, Oslo; 1970.
4. Broch, E., “The development of unlined pressure shafts and tunnels in Norway”. In: Rock mechanics: Caverns
and pressure shafts, published by A.A. Balkema; 1982.
5. Broch, E., “Development of Unlined Pressure Shafts and Tunnels in Norway”, Underground Space, Vol. 8,
No. 3; 1984.
6. Brekke, T.L., Bjorlykke, S. and Blindheim, O.T., “Finite element analysis of the Byrte unlined pressure
shaft failure”. In: Large permanent underground openings. Universitetsforlaget, Oslo, Norway; 1970.

Published in Hydropower & Dams. Issue Three, 2017

8
 The design of unlined hydropower tunnels and shafts:100 years of Norwegian experience

7. Bergh-Christensen, J., “Failure of the unlined pressure tunnel at the Askora” (in Norwegian). Annual national
rock construction conference (Fjellsprengningsteknikk - Bergmekanikk), Trondheim, Norway; 1975.
8. Panthi, K. and Basnet, E.C.B., “Review of the major failure cases of unlined pressure shafts/tunnels of
Norwegian projects”; Hydro Nepal, No. 18, January 2016.
9. Vogt, J.H.L., “Pressure tunnels and geology” (in Norwegian) Report No. 93, Norwegian Geological Survey;
1922.
10. NGI, “Overview of Norwegian unlined tunnels and shafts with water pressures more than 100 m, plus a few
with lower pressures”, Norwegian Geotechnical Institute, Internal Report No. 54402 (in Norwegian); 1972.
11. Garshol, K. and Blindheim, T., “Fosmark hydropower plant. Leakage from pressure shaft” (in Norwegian).
Annual national rock construction conference (Fjellsprengningsteknikk - Bergmekanikk); 1988.
12. Bergh-Christensen, J. and Dannevig, N.T., “Engineering geological evaluations of the unlined pressure shaft
at the Mauranger hydropower plant” (in Norwegian). Unpublished report. Geoteam A/S, Oslo, Norway; 1971.
13. Broch, E., “Design of unlined or concrete lined high pressure tunnels in topographical complicated areas”;
Water Power & Dam Construction; Vol. 36, No.11; 1984.
14. Palmström, A. and Buen, B., “Landslides caused by rupture in hydropower tunnel” (in Norwegian); GEO
Magazine; March 2017.
15. Palmström, A. and Schanche, K., “Design features at Tjodan save time and money”; Water Power & Dam
Construction, June 1987.
16. Dahloe, T.S., Bergh-Christensen, J. and Broch, E., “A review of Norwegian high-pressure concrete plugs”.
Proceedings, Hydropower 92, Norway. Published by A.A. Balkema; 1992.
17. Broch, E., “Sharing experience with concrete plugs in high pressure tunnels”, Hydro Review Worldwide, Vol.
7, No. 4; 1999.
18. Palmström, A. and Stille, H., Chapter 8.5.2 in: Rock Engineering, 2nd Edition; ICE publishing, London, UK;
2015.
19. Hope, J., Palmström, A. and Finnerud, K., “Rebuilding of the 70 years old Nore I power plant”,
Proceedings, Hydropower 97, Trondheim, Norway; 1997.
20. Buen, B. and Palmström, A., “Design and supervision unlined hydropower shafts and tunnels with head up
to 590 meters”, Proceedings, Symposium on Rock mechanics: caverns and pressure shafts. Published by A.A.
Balkema, Netherlands; 1982.
21. Palmström, A., “Norwegian design and construction experience of unlined pressure shafts and tunnels”,
Proceedings, Hydropower 87, Oslo, Norway; 1987.
22. Carter, T.G., Castro, S.O., Carvalho, J.L., Hattersley, D., Wood, K., Barone, F.S., Yuen, C.M.K. and
Giraldo, D., “Tunnelling Issues of Chilean Tertiary Volcaniclastic Rocks”, Proceedings, MIR 2010. XIII
Ciclo di conferenze di Meccanica ed Ingegneria delle Rocce, Torino, Italy; November 2010.

Dr Arild Palmström, a Geological Engineer, is retired from RockMass as, Norway. He has worked on numerous
hydropower projects, road tunnels and subsea tunnel projects in Europe, Asia and South America. He has
contributed more than 50 journal articles, and written two handbooks on rock engineering, as well as the textbook
Rock Engineering (2015)

Consultant, Övre Smestad vei 35e, 0378 Oslo, Norway

Dr Einar Broch retired as a Professor in Geological Engineering from the Norwegian University of Science and
Technology. He is former President of the International Tunnelling Association (1986-89), and former editor of the
journal Tunnelling and Underground Space Technology where he worked from 1986 to 2011 and has been
involved in hydropower projects in 15 countries

Department of Geoscience and Petroleum, Norwegian University of Science and Technology, Sem Sælands vei 1,
N-7491, Norway

Published in Hydropower & Dams. Issue Three, 2017

9